Dr. Lian,

Thanks for always answer my question.

I have an issue with my data, I have a panel data for 8 years.

I'd like to use FE regression, however, How can I check whether there is enough within firm variation for the variable in interest.

For example, I wanna to how does the companies' owner's share affect the firm's performance, However, only around 10% firm in the sample had change in their owner's share.

Is there a systematic ways to check the within varaition in the firm.

I tried to use the command "xtsum", but it give me the b/w S.D. and within S.D..
The within S.D. is around 50% of the b/w S.D. ,  Does this mean the within firm variation is large? If I was wrong, thus How should I use the information.
Thanks